(
sample: Dict,
language: str=None,
mode: str="pretrain",
)
| 47 | |
| 48 | |
| 49 | def process_sample( |
| 50 | sample: Dict, |
| 51 | language: str=None, |
| 52 | mode: str="pretrain", |
| 53 | ) -> Iterable[PromptSample]: |
| 54 | if mode == "pretrain": |
| 55 | prompt = "" |
| 56 | else: |
| 57 | prompt = sample["prompt"] |
| 58 | |
| 59 | try: |
| 60 | if language is not None and language in LANGUAGE_TAG.keys(): |
| 61 | code = LANGUAGE_TAG[language] + "\n" + sample["code"] |
| 62 | else: |
| 63 | code = sample["code"] |
| 64 | except Exception as e: |
| 65 | print(e) |
| 66 | print("The key 'code' is missing in data. Aborted") |
| 67 | exit(0) |
| 68 | |
| 69 | yield PromptSample(prompt, code) |
| 70 | |
| 71 | |
| 72 | def generate_prompt_samples( |
no test coverage detected